Emerging World (Dis)Order
With military confrontation back in full force this week, Washington and Tehran are headed toward another collision course resembling the peak of hostilities last spring. The June ceasefire framework has effectively eroded and been replaced by a familiar resumption of kinetic attacks, primarily targeting the Iranian coastline of the Strait of Hormuz, as both the US and Iran attempt to use the waterways and battlefield to reconfigure the power balance for the next round of negotiations.
As the battle rages, commercial traffic through the Strait of Hormuz has drawn to a halt, with the US-facilitated southern route increasingly viewed as too risky after Iranian missile attacks on commercial vessels there. Iran’s strikes were intended to demonstrate its commitment to challenge any opening of the Strait outside of its direct control and the southern route’s closure will further strain global energy supplies in the near-term.
The US response has focused on degrading the capabilities that allow Iran to exert pressure around Hormuz—coastal missile systems, surveillance infrastructure, and fast-attack assets—while thus far avoiding the kind of civilian targets that would indicate a broader campaign against the Iranian state itself. Tehran’s subsequent retaliation against US positions in Bahrain, Kuwait, and Jordan followed a similar logic, but could readily expand in-kind should the US being to hit Iranian energy or civilian infrastructure.
The proxy dimension has received less attention in recent weeks, but renewed Houthi activity against Saudi Arabia has pushed the issue back into focus. The reported strikes against Abha airport highlight that the boundaries of this confrontation are not entirely controlled by Washington and Tehran. With Iran threatening to call upon the Houthis to interfere in Red Sea shipping should the US expand its campaign, the prospect of proxy action introduces additional escalatory risks to the region and global economy.
Weekly Wildcard
OpenAI this week expanded global availability of its latest advanced AI models after a clash with the Trump administration over access restrictions highlighted the growing geopolitical stakes around artificial intelligence. In June, Washington pushed for tighter controls on the release of leading US models, including restrictions affecting Anthropic, before reversing course. Although temporary, the standoff demonstrated a vulnerability in the US AI ecosystem—access to critical technology can become subject to political decisions, and once that risk is visible, users begin to adapt.
AI ecosystems are difficult to rebuild once users begin adapting to alternatives. Foreign developers that shifted workloads toward open-source or open-weight models during the June uncertainty began rewriting integrations, completing internal approvals, and adjusting compliance processes. Even if US models regain their previous advantage, some users may decide that diversification is worth maintaining.
China is positioned to benefit from the uncertainty created by US technology restrictions. Beijing’s latest data showed weak domestic demand offset by strong exports, with semiconductors, AI hardware, and computing infrastructure driving external growth. In AI software, politically driven pauses in US model availability have created a forcing event for global developers to test Chinese alternatives, many of which are now capable enough for enterprise use and are less prone to domestic political shocks.
At the same time, European governments debating strategic autonomy in defense, energy, and infrastructure face a similar question in AI—whether critical systems should depend on suppliers whose availability could be affected by another government’s political decisions. Sovereign AI initiatives are partly a response to that concern, and the June episode strengthened the argument for reducing reliance on any single provider.
